現今各種開發技術的發展,提升了人們對森林等資源的開發利用,但是隨著採伐速度的加快,森林的覆蓋面積逐年減少中。土地利用會對環境及生態系統造成不同程度的衝擊,若能妥善規劃土地利用,將有助於避免土地濫用造成的影響,除此之外,土地發展規劃的決策常常必須同時滿足許多不同目標。 本研究探討森林採伐規劃問題,此問題為在給定森林區域、各單位的採伐量、未採伐森林土地間的相關性數值以及各目標函數之權重值的條件下,運用三種人工智慧演算法,包含基因演算法(Genetic Algorithm, GA)、免疫演算法(Immune Algorithm, IA)以及粒子群演算法(Particle Swarm Optimization, PSO),規劃出適當的森林採伐區域,使收成量達到最佳化,並提出新的編碼方式求解此問題。測試問題分為兩個部分,第一部分是自設矩陣問題,第二部分是實際的地圖問題。本研究比較此三種演算法對於森林採伐規劃問題的表現,測試數值結果顯示,免疫演算法與基因演算法的求解品質優於粒子群演算法,而粒子群演算法的收斂代數優於其他兩種演算法。
Nowadays, the various development technologies have promoted the utilization of resources such as forests. However, with the speed of harvesting, the area of forest cover is decreasing year by year. Land use will have a different degree of impact on the environment and ecosystems. If land use is properly planned, it will help to decrease the impact of land abuse. In addition, land development planning decisions often have to meet many different objectives at the same time. This study explored the forest harvesting problem. In this forest harvesting problem, given the forest area and relevant values of forest land not harvested and related constraints, we apply three artificial intelligence algorithms, including Genetic Algorithm (GA), Immune Algorithm (IA), Particle Swarm Optimization (PSO), for maximizing the total profit. In this study, GA was used to solve the forest harvesting problem. In this forest harvesting problem, given the forest area and related constraints, the objective is to maximize the total profit. There are two sets of test problems in this thesis, namely, (1) self-designed matrix problems, and (2) practical map problems. We compare the performance of these three algorithms for the forest harvesting problem. Nnumerical results show that IA is superior to the other two algorithms and PSO is faster than the other two algorithms.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。